Use MediaCodec.setOutputSurface() for fullscreen transitions on M.

content/browser/media/android/browser_surface_view_manager.cc

Index: content/browser/media/android/browser_surface_view_manager.cc
diff --git a/content/browser/media/android/browser_surface_view_manager.cc b/content/browser/media/android/browser_surface_view_manager.cc
index 00c51fe1590559dc2a266d69fd6ee75c2b40c573..1a364a4e928d8bbe4cda7969e02210e72b85d058 100644
--- a/content/browser/media/android/browser_surface_view_manager.cc
+++ b/content/browser/media/android/browser_surface_view_manager.cc
@@ -107,9 +107,12 @@ bool BrowserSurfaceViewManager::SendSurfaceID(int surface_id) {
 void BrowserSurfaceViewManager::SendDestroyingVideoSurfaceIfRequired(
     int surface_id) {
   // Only send the surface destruction message on JB, where not doing it can
-  // cause a crash.
-  if (base::android::BuildInfo::GetInstance()->sdk_int() >= 18)
+  // cause a crash, and on Marshmallow+ where we can use this signal to perform
+  // a smoother full screen transition.
+  if (base::android::BuildInfo::GetInstance()->sdk_int() >= 18 &&
+      base::android::BuildInfo::GetInstance()->sdk_int() < 23) {

[watk, 2016/11/03 20:26:34]

We might want to just do this always now. I restricted it initially to see if it
caused crashes. But if it's enabled for a decent % of devices I reckon we should
enable it for all devices because it will let us remove the deferred error logic
we have in AVDA.

Actually, do you know where we can find ANR rates to check how this is going on
JB? I assume that's how deadlocks would manifest. Other kinds of failures should
result in crash reports at least.

[DaleCurtis, 2016/11/04 01:07:22]

Probably you can use crash/blame with the file path to this point and see what
pops up. I can poke around tomorrow.

     return;
+  }
 
   base::WaitableEvent waiter(base::WaitableEvent::ResetPolicy::AUTOMATIC,
                              base::WaitableEvent::InitialState::NOT_SIGNALED);